Market share model for a multi-airport system

Abstract

Because capacity at existing airports was limited and/or because the cost of congestion was becoming unacceptable, several large cities around the world have had to build a second or third major commercial airport to keep up with the demand for air transportation. Such groups of competing airports are called multi-airport systems (M.A.S.) There is extensive historical evidence suggesting that multi-airport systems have often been poorly understood, resulting in disastrous investments such as the construction of airports that remained underused for very long periods of time. The purpose of this paper is to provide a better understanding of the ways M.A.S.s function. First, we consider qualitative characteristics of multi-airport systems, showing the importance of market forces. Then, we build an airport market share model that captures the dynamics of the market, where airlines and air passengers select an airport on the basis of a broad range of factors. Case studies are carried out for several origin-destination markets out of three large metropolitan areas: New York, Washington-Baltimore, and the San Francisco Bay Area. The results show that an airport market share can be well approximated by using few explanatory variables: frequency of service and average fare at the designated airport, and average fare at competing airports. In spite of the relative simplicity of our statistical model, we obtain a good fit between observed and predicted market shares. The explanatory variables are statistically significant and the estimated elasticities (direct price, frequency, and cross-price) are consistent with intuition. We conclude by highlighting the limitations of the model and by suggesting some implications concerning the construction of new airports in metropolitan areas and the potential for regional airports to alleviate the congestion problems at large metropolitan airports.